➡️ READ THE ARTICLE: www.chess.com/blog/FairPlay/c...
➡️ Get My Chess Courses: www.chessly.com/
➡️ Get my best-selling chess book: geni.us/gothamchess
➡️ My book in the UK and Europe: bit.ly/3qFqSf7
➡️ Mein Buch auf Deutsch: bit.ly/45fKt3R
➡️ Mi libro en Español: bit.ly/3Y5xaRx
➡️ Start Playing Chess FOR FREE: bit.ly/3Xa3EsB
➡️ Enjoy my videos? Donate Here : www.paypal.me/gothamchess
Email me your games: gothamletters@gmail.com
Sponsors, Business, Media: gotham@night.co - [DO NOT SEND GAMES HERE]
⭐️ Follow Me If You Are Amazing:
➡️ CAMEO: www.cameo.com/gothamchess
➡️ FACEBOOK: / gothamchessofficial
➡️ SNAP: / levy.rozman
➡️ INSTAGRAM: / gothamchess
➡️ TWITCH: / gothamchess
➡️ TIKTOK: / levyrozman
➡️ TWITTER: / gothamchess
➡️ GOTHAM DISCORD: / discord
➡️ THUMBNAILS BY: / jchessnoob
Danny sending their errand boys (Gotham, Naka) to convince everybody that their site and star competition (Titled Tuesday) is clean and under control ... LOL
True
Yikes!
8 mins dayum
....
Also hai levy
I like to imagine that Levi just appears in places, and that there was no logistics behind him getting there. He goes to bed, and then just suddenly wakes up somewhere new.
This explains a bunch
Levi is the whoopsie guy from mortal kombat only in my youtube and for chess
Or a Quantum Leap kinda thing.
He’s unlocked fast travel
He probably only spawns in when the camera starts rolling tbf
Kramnik frantically trying to find the Infinity Stones to destroy half of Hikaru’s pieces
😂
Lol
Kramnuk, the Mad Giant. Know I got attached to the nickname
Kramnik hasn’t only been mildly reckless. He has completely destroyed his reputation and credibility.
if you're talking about his own reputation, thats true
and thats your opnion after a few tweets ?
No such thing as bad publicity. More people know who he is now.
@@donkarnage6986 throwing accusations of cheating just because you feel like it is truly unpleasent. He accused Hikaru just because his rating is as near as the top engine althought he streams his games.
@@donkarnage6986 it hasn't been just a few tweets lol
The fact that Kramnik didn't even pretend to read the report before being upset with it... everyone needs someone in their life they can trust to tell them when they're losing their mind.
Kramnik automatically doesn’t like anything that was not authored by Kramnik
Right? There's no pleasing the dude. He wanted reports, they made a report, he won't even bother to look at it unless it means his own pedantic requirements. He thinks he's some sort of statistical authority just bc he's good at a board game. Maybe he should put his money where his mouth is and hire some independent statisticians to look at it.
Russians haven't had someone like this for 300 years.
It would be nicer if someone would say "I agree with point x that Kramnik made but disagree with point y". But as always, thinking is hard and hating is easier.
@@junaidahmad1492 ironic then that Kramnik did not do this either, he made his criticisms prior to reading the report and his blog posts this entire time have been extremely dismissive of other people’s work. If one does not engage in reasonable debate and discussion, why would you expect that from others?
When you report someone on CS:GO and it says "report 1541514867146874858 submitted", the 1541514867146874857 previous reports were Kramnik flagging everyone who killed him as a cheater.
For the current state of CS2, 95% of his reports would be accurate.
Except CS:GO doesn't exist anymore, it had a rampant cheating problem, and its successor, CS2, also still has a rampant cheating problem. It's obvious to the eye test, like if you're playing against spinbots etc., but also at higher levels. Just last week, FPL banned their highest rated player for cheating.
@@detrotsidnot FPL (free pro league) it was a faceit ban
The important consideration, which none of the GMs seem to acknowledge, is...how much of the perception of cheating is accounted for by the higher-rated player having a bad day and not wanting to admit it (yes, even GMs)? I mean...I'm 1250 rapid, 1000 blitz and I hammered an 1800 in a blitz tournament earlier in the week. Obviously, he gave me loads of abuse, accused me of cheating and reported me, but...he made a series of blunders that an 800 would've been embarrassed by. Put simply, he was having a crap day, was playing on tilt and took it out on a player 800 points lower-rated.
Totally agree with this. Also factor in someone like me with 150 IQ and an attention deficit, who can go from scatterbrained to world beater and back in a very short amount of time. My rating is some kind of average between those two extremes and depending on when you play me, I might either be amazing or a moron.
Same thing happened to me even with less of a rating difference in a rapid game. I'm sometimes sharp with tactics and I was accused of cheating but I wasn't cheating nor I was banned. Conversely, last year I played 1 game at 98% and I wasn't accused of cheating nor was I cheating either, just played 1 brilliant Vienna game with white.
Yep. Elo rating is not an objective rating where 1 point means you never lose. It’s a way to determine who is more likely to win, and the more games they play the more likely the rating is to be accurate An 1800 can lose to a 1000 but it’s just incredibly unlikely, and as you showed sometimes that happens. If you played him 100 times it might only happen a few times but it will happen statistically
Yeah, there's a couple of other important factors as well. Even if cheating isn't widespread, the *perception* that it is affects players mentally. Not being in a good mental state can really affect performance, causing a player to perform below their potential. Whether it's Levy facing a GM or a GM facing someone they think is cheating (even if they're not), there's a real chance that they perform worse. Thinking that there's cheaters everywhere causes people to start seeing ghosts. But the other factor is that when playing online, it's possible to play a lot of games. Even if cheating isn't widespread, play enough games and you will encounter some. Then things like confirmation bias and availability heuristic will then make it seem like there's a lot more than there are. That sort of mentality is ingrained into the psyche. In the wild, it's better to see a hundred bears that aren't there than to not see one that is. I wonder if people who started out playing online are better able to just brush off games where their opponent cheated and just move on without getting tilted by it.
@@mikeanderson1722bro your iq is not 150 stop the cap😂😂😂
He lied to us, this looks even more like a hostage situation than the previous video.
He said he would have a better set up next trip, not next video
i just want to say thanks to germans for allowing him to still posting this video!❤
------ |
@@sergiob8501Germans? 🧐
@@sergiob8501dude what
Saying kramnik is mildly reckless, is like saying dumbledore spoke calmly to harry during the GoF meme
Levy never fails to have a beef with his barber
Edward Scissorhands?
I have those problems also
😂 👌
I feel for the Jewish Kings out there that hair game is just wild
What are the origins of the never fails bit?
0:40 "Vladimir Kramnik has been mildly reckless" MILDY?!?!????
HMMMMM Nah what you on aboutttttt
But there are streamers you MUST watch, I mean basicaly it is absolutely nessesary to watch them, literally you must do it, watching some streamers, that is what I really mean, some of them are must watch. You just literatuly lost your mind if you dont watch them, I mean it is sad but means you are completely crazy and you dont know what are you talking about when streaming yourself, if you dont watch them. And, another thing,Anish, what all those insuniations like "I dont watch streamers" mean? Can you just come out and tell openly that you hate the most prominent one? Just cone out and say this directly, I know you mean it. Just say this, dont hide under "I dont watch it" bs 🙂
Yeah, that’s a mild understatement
Mildy is an interesting choice of word
Interesting....
14:32 17% refers to score, not winning percentage so its not 'win 1/6 games' as there will be some draws too
It's so levy that he would just dismiss the page about scores then mischaracterise score for the rest of rhe video 😂 7 losses 3 draws from 10 games doesn't seem so wild. Or 15 losses, 1 win, 4 draws from 20.
Omg thank you. Had to scroll a lot for that.
I was looking for this, because that really bothered me too. Hopefully he sees this.
@@frikai8321 why?
@@frikai8321No because the y-axis is still the lower rated player even if they're in the same rating bucket.
4:11 wait, Kramnik hasn’t even READ the report????? jfc
Levy is so lucky that the kidnappers let him record so they can collect the money
What do u mean?
@@LastSamurai21 so u don’t get the lore? Ull understand later
@@LastSamurai21He is in a different place so they make a joke like he's "kidnapped"
A GM can make one bad move; that's all it takes to lose to a 2100 player.
Not, eventually...if you don't see that opportunity and play bad move too, you also can lose
You haven't watched Fabi vs Nepo yet I see...
@@bruce2953 ok, assuming that in one particular game there's drawish situation came after Nepo play ...very not ideal in mid game in important match. That's one game, and what about some genius,that after bad debut and average mid, starts playing like Magnus in endgame? Average probability of average 600, probably 🙂 And they even make strikes 40+ win. And that's just unrated play. Why? Is that Interesting, or not?
Look Magnus games he's always play crap openings in blitz and opponent has advantage but he still outplays not always
@@sHaterred-vh7cs just one little problem. This 600 is not Magnus, definitely ☺️ Maybe Hikaru 🙂🤷 He like strike's
The biggest problem with this report is that underdog victories over the board are often the result of players just being underrated. Being underrated isn’t common online, for obvious reasons.
This ☝️
Kramnik has now officially gone crazy
Just now?
“now”? It’s been a while.
It sneaks up on you if you live long enough. You start to see it around 50 years of age.
@@penknight8532 Biden is over 80 years old, but he's not going crazy like Kramnik. it is unlikely that we will see a crazy Biden at all, since only unworthy people lose their entire reputation in two months
@@penknight8532 Anand is doing just fine probably the highest rated player in his age, he is 54 btw and still is sane
Gotham never fails to title a KZhead video: BANNED For Cheating
BANNED For Cheating.
Technically this just means underdogs don't cheat significantly more than titled players
Other than the super GMs everyone is an underdog to someone. Unless cheaters were the majority it would still result in more upsets not less. Unless you think titled players would cheat more/only against players rated 150-200 pts below them?
“Technically” it means nothing. It’s a set of data. Your extrapolation of some conclusion does not make it “technical”, meaningful, or accurate. “We believe” in a potential interpretation, as this report correctly concludes, is the most you can state. Unless you can explain your technically supported methodology, the validated science behind it, and statistical data to support your conclusion.
I'm not actually making any claims or conclusions with this
I think your point is good. The majority of this report seems to be based on the assumption that players of lower rating will be more likely to cheat than people of higher rating. It doesn't find that correlation, but doesn't test the assumption to make sure it's valid, and instead just sort of assumes that it is when making their conclusion, right?
@@Littlelongy1it literally highlights the possibility its the higher rated players being the ones cheating at the start.... Though it's a real leap to think someone would cheat but only against weaker opponents.... Regardless the whole thing is based on the perception of weaker players cheating to beat stronger players in titled Tuesday ... That is what it's about...
Kramnink’s respect rating went from 2700 to 103 real quick
same with Naka after his Kick stream, tbh
@@SindamscWhat happened?
@@Sindamsc signing to Kick in general or did he say or do something while streaming over there? I’ve never seen a kick stream so very much out of the loop if it’s the latter.
@@Sindamsc whatt stream?
@@shamrock73he had a stream where he was promoting gambling, whilst he was playing slots.
This is actually a really great breakdown... kudos for Levy for taking time to go through this for us even during the vacation
Vacation?? He's being held hostage!!
Kramnik is chronically online 💀
Just as you are
@@Why_did_KZhead_add_handles just as you are
@@ahmed.abdelaleem Just as you are
just as you are
just as you are
Kramnik writes a paragraph about how the report should be conducted. Then he doesn't read it. 🍷🗿
Title for Future reference: “BANNED for Cheating.”
thanks, very helpful (still the same after 16 minutes)
There’s a period at the end
BANNED for Cheating.
Still the same
Interesting story from Minecraft speedrunning: Player Dream was accused and convicted of cheating; it later turned out his accuser knew what to look for because he himself was cheating. Makes me wonder if Kramnik is playing in an ...interesting manner.
One major flaw of this analysis is that it assumes only underdogs cheat. It is possible that the higher rated player cheated and it is possible that both of them cheated. Granted, the underdog winning against the higher rated through cheating is the most problematic scenario, but since we have seen how many times the underdog would win even OTB, it makes perfect sense for a higher rated to cheat as well, esp. if the rating difference is small.
Valid point. But by that standard you have to realize that a for example 2200 is going to only cheat against 2000 and not against higher rated players to make this make sense. That isnt the most logical conclusion so its a occams razor scenario and the most logical conclusion is cheating isnt as big of a problem as it is portrayed online.
If both people cheat, their score will be 50% on average. We don't see that here
@@alexisperron-brault6009that also isn’t really valid or do you think 100% of titled players are cheating?
@@clappedbyben5438 if everyone cheats, everyone will win about 50% of the time. This is not what we see. People might cheat only when it matters though (for one specific game)
@@clappedbyben5438 Dunno if you can call Occam's razor on this, there simply isn't enough information to draw any solid logical conclusions.
When I was 1180 I beat a 1750 in classical... People often "play down" when there is a big difference. They look at a position and think, "Well, he's not strong enough to punish me so I can just do this easy thing and then win and go to lunch" and instead they lose because I'm trying my very best.
There is one other possibility to explain the data: that cheating is so rampant among both the higher and lower rated players in any given online game that the cheating cancels each other out. It is a little like the steroid era in baseball where a lot of pitchers and hitters were both taking performance enhancing drugs. I think that the more definitive approach may be to evaluate how much players performed above/below their rating in OTB vs Online games. If players are consistently playing at a level well above their expected level given their current rating in online games, then that would be an interesting indicator of whether cheating is as prevalent as some would say.
Exactly. Plus that the cheaters are all underdogs is a self defeating assumption. By definition, cheaters cheat to win and gain ratings. So how can they both cheat and stay with very low ratings?
I'm a simple man, I see Gotham, I click
@@_JeffJeff_ I'm a simple man, I see a simple man meme, I want to take part.
Roll*
@@Cage66666 ?
I never fail to be a simple man and see Gotham
Might as well Levy has Good Stuff dam we need another Candidates tournament or other for more recaps they were fun 😅
As someone who does academic business research and statistics all the time: the moment I saw the lack of methodology, correlation diagrams, ANOVA's, p-values, standard errors, and even the text not being symmetrically aligned, I knew that this "study" was not going to be anything to bet your money on.
"This video very interesting" *Block and Report* - Kramnik 2024
Hi Levy, regarding the P-value comment, I think it is relevant for hypothesis testing. You could model the distribution of players as a mix of "noncheating players" and "cheating players" and then run a statistical test. There is much more work to be conducted here. Also, there are better methods than ELO, Glicko might work better idk. Sorry for nerding out 🤓
interesting conclusion at the 2min mark. You could also say: higher rated players are more likely to cheat online, because they have more to lose.
Exactly
My takeaway from this entire video is the FIDE rating system isn't rating players very accurately, and the datasets are not an accurate representation of players' abilities.
Strictly speaking, nothing rates players _very accurately_, all rating systems just look at your results and make some educated-ish guesses. Better sample sizes give better predictions.
"mildly reckless"
Gotham: What many people don't know about me is that I actually have a PhD in Clickbait.
Gotta love how Kramnik got 69 downvotes lol
Levi never fails to make me feel like a mere statistic as he click baits me
17% score doesn't mean that an underdog would win 1/6 of the games. Maybe it's 26% for a draw and 4% for a win, and the result is 0.04×1 + 0.26×0.5 = 0.17.
How long have you been held hostage in Hikaru's ceiling?
Kramnik really said "i ain't reading all that"
After reading the report I am not fully convinced on it. I would have loved to see an F score for the classical FIDE because that looked somewhat distinct. This is also working on the assumption that the lack of distinction in win rates is sufficient, which i'm not fully convinced was sufficiently shown, is a necessary sign. If cheating functions independently of rating, the effect would be totally masked; perhaps higher rated players see a "miracle" move that bails them out vs lower players and can play it off due to their rating, while lower players are more conservative. I think publishing a distribution (without specifics) of found cheater's ratings would be beneficial.
This the reason Indian GM avoid playing online, just play OTB and win everything who cares about title tuesday anyway.
🤡
True that. OTB is and always will be the heartbeat of chess.
I think this report shows the difference in the stress in online versus over the board. The better player tends to win more online because there’s less stress causing variance in skill. Maybe that’s an oversimplification but it stands to reason.
I still laugh whenever I talk about chess with people and they’ll be shocked that you can cheat at chess.
When comparing two groups like this, confidence intervals are way more important than p value which is to compare against the null hypothesis. So, unless you can come up with what should happen by chance, then it's going to be more accurate to use confidence intervals. But yeah there's a missing piece without that.
Since chess is not a game of chance (except whether you play as white or black) I think it's not possible to determine what would happen by chance. You can only look at averages. Maybe the closest bit of information you could get to "what would happen by chance" would be seeing how often strong bots win against themselves as opposed to draw. Since the theory is that chess played perfectly should always end in a draw, so anything else that happens (when players are bots) will be by chance as it were.
My friend, you’re the only person who knows Statistics in the comments section. Everybody else is so uninformed with the basics of 2-sample hypothesis testing 😂. Good job. What is your background in Stats?
@@Odysseusf Haha, kind of. I majored in math so I have some stats. Mostly I got brushed up on it when going into more data science lately.
@@Odysseusf I actually posted a much longer response explaining in more detail, which seems to be taken down. Weird.
@@jennymulholland4319 We say by chance when we mean the null hypothesis which is that no cheating occurred. So if you have a 2200 play a 2500, then they should win 10% of the time by chance (as opposed to skill you could think of it). This is based on rating. Of course rating has to be accurate to know this percentage, which it is admitted that the OTB ratings are not right. To do the p-test you have to stay within on group. So you look at the 2200s vs. 2500s in online only and see if it is more than 10%. The p-value tells you the likelihood that what we saw happen was described by cheating in our case as opposed to "chance." It takes into account how hard it is to achieve and how powerful your sample is. So if you only had one game of a 2200 vs. a 2500, and the 2200 won, you wouldn't claim cheating even though the 2200s won 100% of the games. That could happen by chance. It could be the 10% since it was only one game. But if you had thousands of games and the 2200s won even 30% of the time, that's huge, way bigger than expected if cheating wasn't occurring. Also think if 2200s only typically won 0.05% of the time, then even seeing a single win would indicate cheating because it should be so rare. Anyway, that's what's meant by chance, not that the game has a random element to it. The other problem with this approach is that we are running too many tests. Every comparison is another test, so comparing 2500s to 2400s, 2500s to 2300s, 2500s to 2200s, etc. By chance, one of them is going to indicate cheating has occurred. If you flip a coin a million times, your going to see 100 heads in a row at some point just by chance. When you go digging for results, you're bound to find them even if they aren't there.
Mildly reckless is the understatement of the year...
levy never fails to be held hostage
in my opinion ANYONE that types like THIS, is emotionally unhinged
Kramnik going how Bobby did
Kramnik has not only destroyed his reputation, but has slowly begun to chip away at others’.
as someone taking the AP Statistics exam soon, 21:57 genuinely jumpscared me
The lights look like Levy has cat ears
Levy finally using his stats degree😂😂😂
Levy: _"What are the P Values?"_ Magnus: _"You mean Probability?"_ Hikaru: _"I'm Pushing P!"_ Kramnik: _"It's P as in Psychosis . . ."_ - j q t -
I agree with your take on where are the P-Values and the other parts of a statistical examination that would provide context.
i feel like there's a transmissable thing from cheating in video game speedruns, where a lot of the cheating is actually people who are good enough to do something legit cheating to do it with less time invested. someone who is good at chess might cheat to beat someone who is worse just on feeling entitled to victory without 'wasting their time'
I wonder if at least part of the reason the higher ranked player performs better in titled Tuesday than OTB is just because of facing less pressure and therefore making fewer blunders, errors, etc. It definitely seems a likely explanation
Plus these highly rated players contain more cheaters and engines don’t blunder
Hostage simulator
I really like the vibe of your travel set up. It makes it feel like I’m just watching someone on a FaceTime.
For some odd reason I wanna watch these upset games. The underdog winning that 17% chance.
Levy never fails to be Banned in Chess
The craziest thing in this video is Levy saying Kramnik was "mildly" reckless. Sheesh.
He was trying to be polite methinks.
what has Kramnik been doing and where can I read or watch about it?
@@chunklum3636 Gotham has actually covered it quite extensively on this channel, although i know he doesn't title his videos too accurately so it might be hard to find. "Cheating" / "kramnik" should be in the titles though
@@flookaraz thanks for the help
You know i needed this 3 months ago for my school project
my brain isnt braining
Kramnik: You cheat Hikaru: the 3rd highest rated player🗿
highest performance rating this year
I think an important thing here is also the coverage, we see more titled Tuesday than over the board games
Mildly reckless is an incredible understatement 🤣
kramnik probably the most unlikable person in chess.
Not a chance with Carlsen and Kasparov around... Not to mention Firouzja
@@nuwandalton maybe stupidest take i've ever read
@@JosharooIs that so? Don't be so harsh with yourself. You're stupid, granted, but not THAT stupid. Now run along, fanboy
It used to be hikaru after that I literally don’t care thing. Looks like that keeps changing with time.
No.
beginning pause was so good I thought the video wasn't playing
"traveling?" Everyone knows you enjoy getting kidnapped.
GM Gotham Soon!
Why do you comment so much
@@Hiiigh_Yoshihe probably likes the notification pop up
stop spamming first
(First)²
(First)³
The simple fact of the matter is; if you respect chess, if you really are trying to be your personal best, you would never cheat. To everyone else, shame on you for your actions and you should go cheat elsewhere, not Chess. This is a game for those who really respect the game and intend to uncover our own mental abilities in a world that coddles peoples feelings.
That pic of Levy in that FIDE ranking list got me😂
Gotham already had to cheat to get to gm?? Smh
Kiddo
Gotham - was your barber blind bro? Cause wowwwwwww 😅
A lot of you are seeing the name behind the comment, and not focusing on the comment Kramnik left. While I also believe he goes overboard when assuming people are cheating, he has a very valid point on the actual stats being used. Let's not be quick to dismiss the point being made just because the person making it has made less valid points in the past.
Kramnik will find a way to disagree with literally anything that doesn't match his initial bias, giving any reason he can
Let me guess it’s a Clickbait
Fool watch the video
There are no clickbaits on this channel 😀 There is no war in Ba Sing Se 😐
Levy never fails to cheat.
When you are talking about the rating difference you seem to have forgotten they are using Fide blitz for both ....
these data visualizations are beautiful - I would love to see more of this. maybe even plot some of my own ideas.
It just means higher rated players cheat more than lower rated ones.
Even though they are the same people? They cheat when they are higher rated and stop cheating when rhey are lower rated?
@@DoddyIshamel wut?
@@Prometheus4096 literally what I said. Do you mean they cheat against weaker opponents but not stronger?
@@DoddyIshamel No. I meant literally what I said. You think that higher rated players are just 1 person? And that lower rated people are also just one person? And that these are both the same person?
@Prometheus4096 Every "higher rated person" is the "lower rated person" in other games.... Unless they are Magnus.
Not first
Levys lights in the background make it look like he has cat ears
If we're to assume that the majority of cheating cases in TT are instances of so-called "smart cheating" (a bit of engine use here and there), then I think underdog win percentage is a poor proxy for rates of cheating. I think it's probably equally (if not more) likely that the higher rated player would cheat just to "lock in" a win against a weaker (but still strong enough to be challenging) opponent. It could also be that the weaker player is unlikely to cheat enough to "close the gap" against the stronger player, for fear of getting caught.
Generally valid, but on the point of the higher player cheating you lost me. You realize that basically all of the stronger players are also weaker players to someone else. If they are cheating are they just cheating against lower rated opponents? Does that make sense? Or are you saying more then 70% of all players are just blatantly smart cheating?
First
NPC comment, no one cares
What a tool, finish the video first.
What a tool. Finish the video first.
@@Samurai24200 ok
What a tool. Finish the video first.
Nah, "mildly reckless" is crazy 💀
If underdogs are performing better over the board that probably means over the board blitz ratings are less accurate, which seems reasonable given there are fewer ratings over the board.
Down in amateur world, I got accused of cheating in a bullet game this afternoon. I played h3 to stop the ol’ N-Q pin, he had Bg4 premoved. He lost.
That -6% is saying that the lower rated players 2600-2749 are losing more to players over 2750 and higher in titled tuesday. It means the highest rated players are possibly cheating
You can possibly extrapolate this into 6% of moves are played with assistance, like you said being smart with cheating. A 30 move game may have 1 or 2 moves that were engines.
The major issue with their analysis is the lack of like for like data sets. It means their supposed variable is not controlled. This would be genuinely interesting if they analysed Titled Tuesday results vs a contolled similar sized data set of players playing normally (not in titled tuesday). That way only one variable is being altered. FIDE results are simply not stable enough or large data enough for this kind of analysis. Chess website that published this have an extremely stable data set because of the 1000000s of games played, whereas FIDE played too little in this format and new players affect the results too a far greater degree within that data set.
The problem is: if the evaluation system is different in online vs over the board you end up comparing apples with oranges. And there may be some corrections you can make statistically but you need a very large sample size. Now. on line might be easier to get lots of data but on 1on1 games? Not sure. Btw if I hire a data scientist/analyst and he plots bargraphs with no standard deviation or error I’m gonna fire him on the spot.
what if they make a new rating for prized online tournaments. the players with the biggest gap between their prized rating and normal blitz/rapid etc. rating could get tracked and then you can look up easier if theyre legit by just watching a game of them.
One issue with the study is that FIDE blitz ratings are unreliable indicators of player strength as most players have relatively few rated games in that format. Using that as a basis of comparison is very problematic.
the report also shows that the difference in elo between opponents is smaller in otb, so fewer "upsets" would make sense for that reason.
The problem with these stats are that the FIDE ratings are so inaccurate due to lack of games that they cant really be used as an accurate date set.
It's the only one there us ..
@@DoddyIshamel could check only players with enough recent fide blitz games to have confidence that their rating is accurate
@Flamingcloud083 then the sample size would be much smaller. In any case as pointed out the lack of games for players gives them lower rating which should create more underdog results online. So it would only make the results even more stark.
The addendum said they only used games where players' FIDE blitz ratings were based on at least 50 games and at least one game in the last year. I do wonder how many games this excluded then.
There's also the more obvious lack of pressure on Titled Tuesday compared to a real-world tournament. i/e/ Lower rated players just have less pressure and the top palyers have less pressure too online. "Correlation does not equal causation".
brilliant move (!!) for the "a broken clock is right twice a day" quote